Deep Unrolling of Sparsity-Induced RDO for 3D Point Cloud Attribute Coding

2025-09-10

AI Summary

This work addresses lossy compression of attributes on 3D point clouds with known geometric structure. Methodologically, it proposes an end-to-end differentiable compression framework based on multi-resolution B-spline projection: (i) a sparsity-driven rate-distortion optimization is unrolled into a deep unfolding network, rendering the entire projection process differentiable; (ii) cross-scale prediction and coefficient adaptation are jointly modeled, with ℓ₁-norm regularization enforced to promote coefficient sparsity; and (iii) coarse-to-fine multi-scale reconstruction is supported. The key contribution lies in the first integration of B-spline projection, differentiable sparse optimization, and deep unfolding—achieving geometric awareness, computational efficiency, and reconstruction flexibility. Experiments demonstrate significant rate-distortion gains over state-of-the-art methods, while enabling arbitrary-scale attribute reconstruction and end-to-end joint training.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Learning on the Edge & Model Compression
• Data Mining & Knowledge Management: Data Compression

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Given encoded 3D point cloud geometry available at the decoder, we study the problem of lossy attribute compression in a multi-resolution B-spline projection framework. A target continuous 3D attribute function is first projected onto a sequence of nested subspaces $mathcal{F}^{(p)}_{l_0} subseteq cdots subseteq mathcal{F}^{(p)}_{L}$, where $mathcal{F}^{(p)}_{l}$ is a family of functions spanned by a B-spline basis function of order $p$ at a chosen scale and its integer shifts. The projected low-pass coefficients $F_l^*$ are computed by variable-complexity unrolling of a rate-distortion (RD) optimization algorithm into a feed-forward network, where the rate term is the sparsity-promoting $ell_1$-norm. Thus, the projection operation is end-to-end differentiable. For a chosen coarse-to-fine predictor, the coefficients are then adjusted to account for the prediction from a lower-resolution to a higher-resolution, which is also optimized in a data-driven manner.

Problem

• Lossy attribute compression for 3D point clouds
• Multi-resolution B-spline projection framework optimization
• Differentiable rate-distortion optimization with sparsity promotion

Innovation

• Multi-resolution B-spline projection framework
• Unrolling sparsity-induced RDO into network
• Differentiable end-to-end projection operation